Dynamics of the dead wood carbon pool in northwestern Russian boreal forests

Abstract

Our study examines dead wood dynamics in a series of permanent plots established in closed, productive second-growth forest stands of north-west Russia and in temporary plots that represent different successional stages and types of disturbance. Dead wood stores measured on 63 plots 0.2–1.0 ha in size range from 1–8 Mg C ha−1 in young to mature intensively managed stands, 17 Mg C ha−1 in an old-growth forest, 20 Mg C ha−1 on a clear-cut, and 21–39 Mg C ha−1 following a severe windthrow. A total of 122 logs, snags, and stumps aged by long-term plot records was sampled for decay rates and to develop a system of decay classes. Annual decomposition rates are: 3.3% for pine, 3.4% for spruce, and 4.5% for birch. Based on these decay rates the average residence time of carbon (C) in the dead wood pool is 22–30 years. The mortality input on the permanent plots was 23–60 Mg C ha−1 over 60 years of observation or 15–50% of the total biomass increment. This data suggests a dead wood mass of 10–22 Mg C ha−1 would be expected in these mature forests if salvage had not occurred. In old-growth forests, dead wood comprised about 20% of the total wood mass, a proportion quite similar to the larger, more productive forests of the Pacific Northwest (USA). If this proportioning is characteristic of cool conifer forests it would be useful to estimate potential dead wood mass for old-growth forests without dead wood inventories. However, the use of a single live/dead wood ratio across the range of successional stages, a common practice in C budget calculations, may substantially over- or under-estimate the dead wood C pool depending upon the type of disturbance regime. Intensive forest management including short harvest rotations, thinning and wood salvage reduces dead wood C stores to 5–40% of the potential level found in undisturbed old-growth forest. In contrast, natural disturbance increases dead wood C pool by a factor of 2–4.